Mind.1.0.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import sklearn.cluster as cluster\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import StandardScaler as sc\n",
    "from sklearn.neighbors import KNeighborsRegressor as KNN\n",
    "\n",
    "big_data = pd.read_csv('airbnb_coded.2.0.csv')\n",
    "sc = sc()\n",
    "knn = KNN(n_neighbors=65, weights='distance')\n",
    "\n",
    "pd.options.mode.chained_assignment = None  # default='warn'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "26\n",
      "[[  1.   0.   0. ...  97.   1. 100.]\n",
      " [  1.   0.   0. ...  75.   1. 100.]\n",
      " [  1.   0.   0. ...  80.   1.   0.]\n",
      " ...\n",
      " [  1.   0.   0. ...  99.   1.   0.]\n",
      " [  1.   0.   0. ...  90.   1.   0.]\n",
      " [  1.   0.   0. ... 100.   1. 100.]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "\n",
    "ct = ColumnTransformer(transformers=[('encoder', OneHotEncoder(), [-1])], remainder='passthrough')\n",
    "var = 'bathrooms'\n",
    "var2 = 'accommodates'\n",
    "var3 = 'cleaning_fee'\n",
    "var4 = 'bedrooms'\n",
    "var5 = 'beds'\n",
    "var6 = 'minimum_nights_x'\n",
    "var7 = 'maximum_nights'\n",
    "var8 = 'number_of_reviews_y'\n",
    "var9 = 'listing_class'\n",
    "var10 = 'price_x'\n",
    "var11 = 'bed_code'\n",
    "var12 = 'cancellation_policy_mapped'\n",
    "var13 = 'room_code'\n",
    "var14 = 'superhost'\n",
    "var15 = 'instant_book'\n",
    "var16 = 'review_scores_location'\n",
    "var17 = 'review_scores_rating'\n",
    "var18 = 'host_total_listings_count'\n",
    "var19 = 'availability_90'\n",
    "var20 = 'room_code'\n",
    "var21 = 'extra_people'\n",
    "var22 = 'property_type'\n",
    "var23 = 'review_scores_value'\n",
    "var24 = 'review_scores_cleanliness'\n",
    "var25 = 'review_scores_checkin'\n",
    "var26 = 'review_scores_communication'\n",
    "var27 = 'review_scores_value'\n",
    "var28 = 'guests_included'\n",
    "var29 = 'security_deposit'\n",
    "var30 = 'host_since'\n",
    "var31 = 'loc_class'\n",
    "var32 = 'profile_pic'\n",
    "var33 = 'amenities'\n",
    "\n",
    "big_data = big_data[big_data.price_x < 400]\n",
    "big_data = big_data[big_data.price_x > 8]\n",
    "\n",
    "data = big_data[[var, var2, var3, var4, var5, var6, var7, var8, var11, var12, var13, var14, var15, var16, var17, var18, var19, var21, var22, var23, var24,\n",
    "                 var25, var26, var27, var28, var29, var30, var31, var32, var33, var9, var10]]\n",
    "\n",
    "data[var] = data[var].fillna(data[var].median())\n",
    "data[var2] = data[var2].fillna(data[var2].median())\n",
    "data[var3] = data[var3].fillna(data[var3].median())\n",
    "data[var4] = data[var4].fillna(data[var4].median())\n",
    "data[var5] = data[var5].fillna(data[var5].median())\n",
    "\n",
    "data[var6][17990] = data[var6].median()\n",
    "data[var6][6215] = data[var6].median()\n",
    "data[var6][11299] = data[var6].median()\n",
    "data[var6][12737] = data[var6].median()\n",
    "data[var6][14598] = data[var6].median()\n",
    "# data[var7][872] = 8000\n",
    "# data[var7][2372] = 8000\n",
    "# data[var7][4391] = 8000\n",
    "# data[var7][17952] = 8000\n",
    "# data[var7][17990] = data[var7].median()\n",
    "# data[var7][6351] = data[var7].median()\n",
    "# data[var7][8341] = data[var7].median()\n",
    "\n",
    "\n",
    "data[var6] = data[var6].fillna(data[var6].median())\n",
    "data[var7] = data[var7].fillna(data[var7].median())\n",
    "data[var8] = data[var8].fillna(data[var8].median())\n",
    "data[var10] = data[var10].fillna(data[var10].median())\n",
    "data[var11] = data[var11].fillna(data[var11].median())\n",
    "data[var12] = data[var12].fillna(data[var12].median())\n",
    "data[var13] = data[var13].fillna(data[var13].median())\n",
    "data[var14] = data[var14].fillna(data[var14].median())\n",
    "data[var15] = data[var15].fillna(data[var15].median())\n",
    "data[var16] = data[var16].fillna(data[var16].median())\n",
    "data[var17] = data[var17].fillna(data[var17].median())\n",
    "data[var18] = data[var18].fillna(data[var18].median())\n",
    "data[var19] = data[var19].fillna(data[var19].median())\n",
    "data[var20] = data[var20].fillna(data[var20].median())\n",
    "data[var21] = data[var21].fillna(data[var21].median())\n",
    "data[var22] = data[var22].fillna('Apartment')\n",
    "data[var23] = data[var23].fillna(data[var23].median())\n",
    "data[var24] = data[var24].fillna(data[var24].median())\n",
    "data[var25] = data[var25].fillna(data[var25].median())\n",
    "data[var26] = data[var26].fillna(data[var26].median())\n",
    "data[var27] = data[var27].fillna(data[var27].median())\n",
    "data[var28] = data[var28].fillna(data[var28].median())\n",
    "data[var29] = data[var29].fillna(data[var29].median())\n",
    "data[var30] = data[var30].fillna(data[var30].median())\n",
    "data[var31] = data[var31].fillna(data[var31].median())\n",
    "data[var32] = data[var32].fillna(data[var32].median())\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# data[var6][17990] = data[var6].median()\n",
    "\n",
    "# data = data.drop(data.index[17990])\n",
    "data = data.drop(data.index[14598])\n",
    "\n",
    "data = data[[var2, var3, var6, var18, var19, var21, var8, var16, var17, var, var29,\n",
    "             var9, var13, var12, var31, var10]]\n",
    "# data = data[[var, var2, var3, var6, var8, var18, var17, var21, var19, var9, var12, var20, var22, var10]]\n",
    "\n",
    "\n",
    "X = data.iloc[:, :-1].values\n",
    "y = data.iloc[:, -1].values\n",
    "X = ct.fit_transform(X)\n",
    "# X = X.todense()\n",
    "X = np.array(X)\n",
    "X = ct.fit_transform(X)\n",
    "# X = X.todense()\n",
    "X = np.array(X)\n",
    "X = ct.fit_transform(X)\n",
    "# X = X.todense()\n",
    "X = np.array(X)\n",
    "X = ct.fit_transform(X)\n",
    "# X = X.todense()\n",
    "X = np.array(X)\n",
    "# X = ct.fit_transform(X)\n",
    "# # X = X.todense()\n",
    "# X = np.array(X)\n",
    "# X = ct.fit_transform(X)\n",
    "# # X = X.todense()\n",
    "# X = np.array(X)\n",
    "# X = ct.fit_transform(X)\n",
    "# # X = X.todense()\n",
    "# X = np.array(X)\n",
    "\n",
    "# from sklearn.feature_extraction.text import CountVectorizer\n",
    "# vect = CountVectorizer()\n",
    "# amenities = data[var33].values\n",
    "# # print(amenities)\n",
    "# # for i in range(len(amenities)):\n",
    "# # print(amenities)\n",
    "# vect.fit(amenities)\n",
    "# amenities_array = vect.transform(amenities)\n",
    "# amenities_array = amenities_array.todense()\n",
    "\n",
    "# X = np.concatenate((X, amenities_array),axis=1)\n",
    "# plt.scatter(data[var6].values, data[var7].values, s=100, c='red', alpha=0.5)\n",
    "# plt.show()\n",
    "\n",
    "print(len(X[0]))\n",
    "print(X)\n",
    "# X[:,-8:] = sc.fit_transform(X[:,-8:])\n",
    "# X = sc.fit_transform(X)\n",
    "# print(X[0])\n",
    "# X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.1)\n",
    "\n",
    "# data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "metadata": {},
   "outputs": [],
   "source": [
    "import keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "metadata": {},
   "outputs": [],
   "source": [
    "ann = Sequential()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-305-be550337f789>:1: UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(activation=\"relu\", input_dim=26, units=17, kernel_initializer=\"uniform\")`\n",
      "  ann.add(Dense(output_dim=17, init='uniform', activation='relu', input_dim=26))\n"
     ]
    }
   ],
   "source": [
    "ann.add(Dense(output_dim=17, init='uniform', activation='relu', input_dim=26))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-306-5cb98ab76bde>:1: UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(activation=\"relu\", units=17, kernel_initializer=\"uniform\")`\n",
      "  ann.add(Dense(output_dim=17, init='uniform', activation='relu'))\n"
     ]
    }
   ],
   "source": [
    "ann.add(Dense(output_dim=17, init='uniform', activation='relu'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-307-5cb98ab76bde>:1: UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(activation=\"relu\", units=17, kernel_initializer=\"uniform\")`\n",
      "  ann.add(Dense(output_dim=17, init='uniform', activation='relu'))\n"
     ]
    }
   ],
   "source": [
    "ann.add(Dense(output_dim=17, init='uniform', activation='relu'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-308-5cb98ab76bde>:1: UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(activation=\"relu\", units=17, kernel_initializer=\"uniform\")`\n",
      "  ann.add(Dense(output_dim=17, init='uniform', activation='relu'))\n"
     ]
    }
   ],
   "source": [
    "ann.add(Dense(output_dim=17, init='uniform', activation='relu'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-309-5cb98ab76bde>:1: UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(activation=\"relu\", units=17, kernel_initializer=\"uniform\")`\n",
      "  ann.add(Dense(output_dim=17, init='uniform', activation='relu'))\n"
     ]
    }
   ],
   "source": [
    "ann.add(Dense(output_dim=17, init='uniform', activation='relu'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-310-de58dc7503ba>:1: UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(units=1, kernel_initializer=\"uniform\")`\n",
      "  ann.add(Dense(output_dim=1, init='uniform'))\n"
     ]
    }
   ],
   "source": [
    "ann.add(Dense(output_dim=1, init='uniform'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "metadata": {},
   "outputs": [],
   "source": [
    "ann.compile(optimizer='adam', loss='mean_squared_error', metrics=['mean_squared_error'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 312,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-312-b1ce75173cec>:1: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
      "  ann.fit(X_train, y_train, batch_size=12, nb_epoch=100)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "12551/12551 [==============================] - 1s 93us/step - loss: 1587.7895 - mean_squared_error: 1587.7903\n",
      "Epoch 2/100\n",
      "12551/12551 [==============================] - 1s 79us/step - loss: 1111.5868 - mean_squared_error: 1111.5863\n",
      "Epoch 3/100\n",
      "12551/12551 [==============================] - 1s 79us/step - loss: 991.4241 - mean_squared_error: 991.4236\n",
      "Epoch 4/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 964.7866 - mean_squared_error: 964.7866\n",
      "Epoch 5/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 934.8412 - mean_squared_error: 934.8417\n",
      "Epoch 6/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 929.3266 - mean_squared_error: 929.3267\n",
      "Epoch 7/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 926.4652 - mean_squared_error: 926.4648\n",
      "Epoch 8/100\n",
      "12551/12551 [==============================] - 1s 83us/step - loss: 915.2774 - mean_squared_error: 915.2763\n",
      "Epoch 9/100\n",
      "12551/12551 [==============================] - 1s 79us/step - loss: 893.2748 - mean_squared_error: 893.2749\n",
      "Epoch 10/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 882.3338 - mean_squared_error: 882.3341\n",
      "Epoch 11/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 869.5278 - mean_squared_error: 869.5273\n",
      "Epoch 12/100\n",
      "12551/12551 [==============================] - 1s 80us/step - loss: 870.5855 - mean_squared_error: 870.5854\n",
      "Epoch 13/100\n",
      "12551/12551 [==============================] - 1s 81us/step - loss: 853.7192 - mean_squared_error: 853.7191\n",
      "Epoch 14/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 843.3884 - mean_squared_error: 843.3887\n",
      "Epoch 15/100\n",
      "12551/12551 [==============================] - 1s 89us/step - loss: 850.1736 - mean_squared_error: 850.1738\n",
      "Epoch 16/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 842.4404 - mean_squared_error: 842.4407\n",
      "Epoch 17/100\n",
      "12551/12551 [==============================] - 1s 79us/step - loss: 839.0158 - mean_squared_error: 839.0154\n",
      "Epoch 18/100\n",
      "12551/12551 [==============================] - 1s 82us/step - loss: 840.8568 - mean_squared_error: 840.8559\n",
      "Epoch 19/100\n",
      "12551/12551 [==============================] - 1s 85us/step - loss: 845.9323 - mean_squared_error: 845.9327\n",
      "Epoch 20/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 829.9166 - mean_squared_error: 829.9163\n",
      "Epoch 21/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 828.8444 - mean_squared_error: 828.8445\n",
      "Epoch 22/100\n",
      "12551/12551 [==============================] - 1s 79us/step - loss: 834.6645 - mean_squared_error: 834.6646\n",
      "Epoch 23/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 821.8709 - mean_squared_error: 821.8712\n",
      "Epoch 24/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 816.3824 - mean_squared_error: 816.3829\n",
      "Epoch 25/100\n",
      "12551/12551 [==============================] - 1s 84us/step - loss: 829.7771 - mean_squared_error: 829.7762\n",
      "Epoch 26/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 809.9069 - mean_squared_error: 809.9073\n",
      "Epoch 27/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 824.6226 - mean_squared_error: 824.6217\n",
      "Epoch 28/100\n",
      "12551/12551 [==============================] - 1s 82us/step - loss: 802.7570 - mean_squared_error: 802.7571\n",
      "Epoch 29/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 844.5676 - mean_squared_error: 844.5680\n",
      "Epoch 30/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 808.2295 - mean_squared_error: 808.2301\n",
      "Epoch 31/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 814.5914 - mean_squared_error: 814.5912\n",
      "Epoch 32/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 805.6174 - mean_squared_error: 805.6176\n",
      "Epoch 33/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 801.3761 - mean_squared_error: 801.3760\n",
      "Epoch 34/100\n",
      "12551/12551 [==============================] - 1s 89us/step - loss: 816.5285 - mean_squared_error: 816.5286\n",
      "Epoch 35/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 796.1494 - mean_squared_error: 796.1495\n",
      "Epoch 36/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 796.9122 - mean_squared_error: 796.9123\n",
      "Epoch 37/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 799.3033 - mean_squared_error: 799.3036\n",
      "Epoch 38/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 809.0977 - mean_squared_error: 809.0981\n",
      "Epoch 39/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 801.7863 - mean_squared_error: 801.7864\n",
      "Epoch 40/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 792.7584 - mean_squared_error: 792.7578\n",
      "Epoch 41/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 790.2732 - mean_squared_error: 790.2729\n",
      "Epoch 42/100\n",
      "12551/12551 [==============================] - 1s 87us/step - loss: 793.7634 - mean_squared_error: 793.7631\n",
      "Epoch 43/100\n",
      "12551/12551 [==============================] - 1s 87us/step - loss: 786.0691 - mean_squared_error: 786.0695\n",
      "Epoch 44/100\n",
      "12551/12551 [==============================] - 1s 81us/step - loss: 790.8139 - mean_squared_error: 790.8140\n",
      "Epoch 45/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 789.6633 - mean_squared_error: 789.6639\n",
      "Epoch 46/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 789.5847 - mean_squared_error: 789.5849\n",
      "Epoch 47/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 783.6863 - mean_squared_error: 783.6865\n",
      "Epoch 48/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 790.9934 - mean_squared_error: 790.9932\n",
      "Epoch 49/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 783.1194 - mean_squared_error: 783.1194\n",
      "Epoch 50/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 780.9792 - mean_squared_error: 780.9791\n",
      "Epoch 51/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 780.0296 - mean_squared_error: 780.0298\n",
      "Epoch 52/100\n",
      "12551/12551 [==============================] - 1s 82us/step - loss: 767.6460 - mean_squared_error: 767.6465\n",
      "Epoch 53/100\n",
      "12551/12551 [==============================] - 1s 108us/step - loss: 773.8341 - mean_squared_error: 773.8339\n",
      "Epoch 54/100\n",
      "12551/12551 [==============================] - 1s 80us/step - loss: 774.5017 - mean_squared_error: 774.5014\n",
      "Epoch 55/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 771.2182 - mean_squared_error: 771.2184\n",
      "Epoch 56/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 777.2541 - mean_squared_error: 777.2538\n",
      "Epoch 57/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 768.4794 - mean_squared_error: 768.4793\n",
      "Epoch 58/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 770.9495 - mean_squared_error: 770.9497\n",
      "Epoch 59/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 767.7814 - mean_squared_error: 767.7813\n",
      "Epoch 60/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 776.5761 - mean_squared_error: 776.5766\n",
      "Epoch 61/100\n",
      "12551/12551 [==============================] - 1s 82us/step - loss: 775.0367 - mean_squared_error: 775.0367\n",
      "Epoch 62/100\n",
      "12551/12551 [==============================] - 1s 86us/step - loss: 775.0988 - mean_squared_error: 775.0991\n",
      "Epoch 63/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 773.5373 - mean_squared_error: 773.5369\n",
      "Epoch 64/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 770.6882 - mean_squared_error: 770.6879\n",
      "Epoch 65/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 764.1184 - mean_squared_error: 764.1185\n",
      "Epoch 66/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 764.8102 - mean_squared_error: 764.8102\n",
      "Epoch 67/100\n",
      "12551/12551 [==============================] - 1s 81us/step - loss: 754.1919 - mean_squared_error: 754.1920\n",
      "Epoch 68/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 763.6051 - mean_squared_error: 763.6050\n",
      "Epoch 69/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 758.3475 - mean_squared_error: 758.3475\n",
      "Epoch 70/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 764.4955 - mean_squared_error: 764.4954\n",
      "Epoch 71/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 764.9322 - mean_squared_error: 764.9322\n",
      "Epoch 72/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 753.0749 - mean_squared_error: 753.0754\n",
      "Epoch 73/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 760.1460 - mean_squared_error: 760.1455\n",
      "Epoch 74/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 761.1094 - mean_squared_error: 761.1095\n",
      "Epoch 75/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 765.7456 - mean_squared_error: 765.7462\n",
      "Epoch 76/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 758.7223 - mean_squared_error: 758.7222\n",
      "Epoch 77/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 752.5760 - mean_squared_error: 752.5759\n",
      "Epoch 78/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 749.3768 - mean_squared_error: 749.3771\n",
      "Epoch 79/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 759.5018 - mean_squared_error: 759.5016\n",
      "Epoch 80/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 758.2107 - mean_squared_error: 758.2106\n",
      "Epoch 81/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 737.2871 - mean_squared_error: 737.2869\n",
      "Epoch 82/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 762.8595 - mean_squared_error: 762.8598\n",
      "Epoch 83/100\n",
      "12551/12551 [==============================] - 1s 74us/step - loss: 749.8028 - mean_squared_error: 749.8018\n",
      "Epoch 84/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 750.4732 - mean_squared_error: 750.4732\n",
      "Epoch 85/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 744.1960 - mean_squared_error: 744.1963\n",
      "Epoch 86/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 744.3459 - mean_squared_error: 744.3459\n",
      "Epoch 87/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 743.6194 - mean_squared_error: 743.6194\n",
      "Epoch 88/100\n",
      "12551/12551 [==============================] - 1s 77us/step - loss: 750.7359 - mean_squared_error: 750.7360\n",
      "Epoch 89/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 732.0182 - mean_squared_error: 732.0178\n",
      "Epoch 90/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 744.0023 - mean_squared_error: 744.0026\n",
      "Epoch 91/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 748.6096 - mean_squared_error: 748.6092\n",
      "Epoch 92/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 735.1267 - mean_squared_error: 735.1265\n",
      "Epoch 93/100\n",
      "12551/12551 [==============================] - 1s 78us/step - loss: 738.3992 - mean_squared_error: 738.3994\n",
      "Epoch 94/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 742.3900 - mean_squared_error: 742.3900\n",
      "Epoch 95/100\n",
      "12551/12551 [==============================] - 1s 74us/step - loss: 745.0211 - mean_squared_error: 745.0211\n",
      "Epoch 96/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 727.8862 - mean_squared_error: 727.8862\n",
      "Epoch 97/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 741.6037 - mean_squared_error: 741.6044\n",
      "Epoch 98/100\n",
      "12551/12551 [==============================] - 1s 76us/step - loss: 731.5992 - mean_squared_error: 731.5993\n",
      "Epoch 99/100\n",
      "12551/12551 [==============================] - 1s 75us/step - loss: 734.3077 - mean_squared_error: 734.3078\n",
      "Epoch 100/100\n",
      "12551/12551 [==============================] - 2s 120us/step - loss: 737.5787 - mean_squared_error: 737.5783\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x7f972b02ad60>"
      ]
     },
     "execution_count": 312,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ann.fit(X_train, y_train, batch_size=12, nb_epoch=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 313,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "761.4802231064082"
      ]
     },
     "execution_count": 313,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_squared_error as mse\n",
    "mse(y_test, ann.predict(X_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 315,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.title(\"ANN Performance\")\n",
    "plt.ylabel('Prediction')\n",
    "plt.xlabel('Real')\n",
    "plt.scatter(y_test, ann.predict(X_test), c='blue', alpha=0.1)\n",
    "plt.plot([1,10,100,200,300,400],[1,10,100,200,300,400],c='red')\n",
    "plt.savefig('ann.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}